I'll start off by saying, I've searched. I have also searched, but I cant seem to find any thread that includes a SOLUTION to this problem.

My enigine was recently rebuilt, and the the shop which did the install did have to cut and splice here and there, due to the age and condition of some of the connectors on the harness. Despite this however, all of the test on the wires and connections have been good.

After my research on the forum, I was convinced it was not the sensor, but a problem with the wiring. They did have the KS and PS switched initially, but we caught that and fixed it. We reset the ECU and still got the code. We re-wired the KS with new sheilded wire...code. We tested the wires and connections again, and everything was good.

We finally tried a used (working(?)) sensor from someone local, and now the code shows up only upon startup, and only until the car warms up, then it goes away.

Any ideas as to what would cause this? Does it affect the way the car runs? I have read some posts tha say the KS code causes limp mode, others say it won't. Which is it, and if its in limp mode when the CEL is on, what happens once it goes off?

Thanks in advance for any help or guidance.

 

According to the workshop manual it retards the timing (=less power and less knock)
This only goes if you are using the stock ECU, the PFC for example doesn't use the knock sensor.


I would guess something concerning the grounding/wiring shorts or opens when shrinking during cool down. But that's just an inexact guess.

 

Maybe during startup when cold, engine noise is enough louder than when warm that pulses are generated from the sensor. Pretty sure the knock sensor is a piezo-electric sensor that acts like a microphone. It seems to be ac-coupled, i.e. an ohmmeter shows infinite resistance to ground even when tapping the block next to it, also no readable voltage on either ac or dc settings on a digital voltmeter. The voltage pulses must be so narrow that a meter can't track them.

I was never able to verify knock sensor operation with a meter, even on a new unit out of the car; maybe it needs an oscilloscope to see pulses.

If any of the brown "goo" has leaked out of the sensor, it will become more sensitive to noise.

 

Shouldn't that be the other way around?

The rest of the things you say verifies in another thread I read on the forum. I think the pulses were measurable with an high-quality sensitive voltmeter (ac). I will soon try if my voltmeter is sensitive enough An osciloscope would be good though.